In my post on Ancient Beer, one of my main sources came over to comment, Merryn and Graham Dinely. Presumably, something pops up when someone links to their site. I should probably have figured that out years ago. Graham suggested that I look into the work of Michael Baillie, a tree-ring expert (dendochronologist, so now you know) whose work indicates terrible conditions 2354-2354 BC, likely resulting in widespread famine. His book is reviewed here at New Scientist. He identifies other dire periods and believes that showers of comets are the most likely explanation. There is no record of volcanoes, and those would not have effects lasting nearly a decade in any event. That particular set of dates is important, as it coincides with Corded Ware and Bell Beaker cultures and the full replacement of Britain's Neolithic farming population with those peoples who are deeply connected with the Yamnaya, the Indo-European nomadic people. Baillie believes that the replacement was not so much the result of invading horsemen with bronze weapons and cattle but of the nine-year famine wiping out the sedentary population across Europe that was crop dependent, while the (relative) newcomers survived better, though likely not entirely well.
It's a highly plausible story, and similar to the new understandings we are developing of disease wiping out a population, or at least weakening it enough that it can be more easily overcome. The most dramatic example is the discovery of the New World and the massive die-offs of the native populations here, but there's no reason why it could not have happened many other times. I would like to see more evidence before I accept the theory, but I am kindly disposed to it at present. Nine years of famine would certainly be a lot for any population to endure. I do immediately wonder about those societies that survived by fishing, though. They would also have a better chance of hanging on, wouldn't they?